Oh man, what the fuck. What I really don't like about this is that you assume some sort of motivation when someone submits a PR. When they provide a feature implementation, a bug fix or a configuration change, you assume that the way things were before did not work for the contributor. That it was so problematic, that they invested quite a bit of time to scratch their own itch.

You simply can't look at a PR in complete isolation and just evaluate whether its changes are positive or not. On some level, you'll decide that, sure, if this user is better off with this change and it doesn't break anything, then might as well merge it.
But there is no user here. It's going to include pointless changes that just need to look plausible enough, so that this bot can bolster its credibility.